forensic document examiner using graphology science

42
Forensics Document Examiner – using graphology science(FODEX) Group Members SHWETA HEGADE (B80364283) GARGEE HIRAY (B80364285) PRAJKTA MALI (B80364335) AISHWARYA SHINDE (B80364400) [email protected] STES’S SMT. KASHIBAINAVALECOLLEGE OF ENGINEERING SAVITRIBAI PHULE PUNE UNIVERSITY Under the guidance of Prof.Punam Raskar

Transcript of forensic document examiner using graphology science

Page 1: forensic document examiner using graphology science

Forensics Document Examiner – using graphology science(FODEX)

Group MembersSHWETA HEGADE (B80364283)

GARGEE HIRAY (B80364285)

PRAJKTA MALI (B80364335)

AISHWARYA SHINDE (B80364400)

[email protected]

 STES’S SMT. KASHIBAINAVALECOLLEGE OF ENGINEERING  SAVITRIBAI PHULE PUNE UNIVERSITY

Under the guidance ofProf.Punam Raskar

Page 2: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Overview• Problem Statement• About Graphology • System Working • Feasibility assessment • Basic Workflow Diagram• Algorithms• UML Diagrams • Mathematical Model • FODEX and Other systems• Implementation till date • Project Plan• Bibliography

Page 3: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

SPONSORED BY

Centre for Personality Assessment and Graphology• They provide Training and Services in

Handwriting, Signature, Colour, Logo Analysis for recruitment, employee assessment, personality assessment, counseling, business consultation, forensic document examination.

• http://www.cpag.in/• https://www.facebook.com/CPAGP

Page 4: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Problem Statement

• Developing a system which will examine document and generate a algorithmic analysis of human handwriting by comparison of different samples of handwriting, through image processing algorithms .

1

Page 5: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

About Graphology

•History•Meaning of Graphology•How graphology works?•What can be revealed?

2

Page 6: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

What do handwriting analyst checks in handwriting ?- LOOPS,MARGINS,BASELINE,SIZE,SPACING ,PRESSURE,SLANTS, ZONES, CONNECTIONS, PASTOCITY,TRAIT METHOD

Example:-

3

Page 7: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Page 8: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

•Forensics and graphology •How our system works?

Page 9: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Feasibility assessment

P- Complete

NP- Complete

NP- Hard

4

Page 10: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Workflow

• Image preprocessing• Thinning• Segmentation• Feature extraction• Profile generation

5

Page 11: forensic document examiner using graphology science

Work Flow

SKNCOE, Department of Computer Engineering 2014-15 6

Page 12: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Basics of Image Processing

•Image•Pixel•Types of Images

–Vector–Digital

•Binary•Grayscale•Color

Page 13: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

RGB splitting

•Color images are often built of several stacked color channels, each of them representing value levels of the given channel. For example, RGB images are composed of three independent channels for red, green and blue primary color components.

7

Page 14: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Grayscale

•Grayscale images are distinct from one-bit bi-tonal black-and-white images, which in the context of computer imaging are images with only the two colors, black, and white (also called bi-level or binary images). The value of each pixel is a single sample.

•Grayscale images have 256of gray .

8

Page 15: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

continue•How all grayscale algorithms fundamentally work

•All grayscale algorithms utilize the same basic three-step process:

1. Get the red, green, and blue values of a pixel

2. Convert these pixel values into gray value.

3. Replace the original red, green, and blue values with the new gray value

•Basic Formula:

Gray = (Red + Green + Blue) / 3

( (0.3 * R) + (0.59 * G) + (0.11 * B) )9

Page 16: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

10

Blurring/Softening

Gaussian Blur salt and pepper

Noise

Page 17: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Threshold Detection

•Thresholding is a non-linear operation that converts a gray-scale image into a binary image where the two levels are assigned to pixels that are below or above the specified threshold value.

•Thresholding Methods:–Histogram shape-based methods.

•E.g. :Ostu’s Method ,Kapur’s Method.–Clustering-based methods

11

Page 18: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

12

Thresholding of An image

Fig.[A ] Fig.[b]

Page 19: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Thinning

•Thinning algorithm is a Morphological operation that is used to remove selected foreground pixels from binary images.

•Thinning Methods:–Iterative Thinning

•e.g.: Stentiford Thinning Algorithm, Zhang-Suen Thinning Algorithm,etc.

–Non-Iterative Thinning•e.g.: Neusius-Olszewski Algorithm

13

Page 20: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

14

Segmentation

•Separation or Isolation of Image of Handwritten Document Text into

–Lines–Words–Characters

•Methods of Segmentation –Histogram based–Profile and contour based–Background analysis

Page 21: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

15

•Segmentation of Image of Handwritten Text.

Page 22: forensic document examiner using graphology science

•Segmentation of Image of Handwritten Text.

Page 23: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

16

•Segmentation of Image of Handwritten Text.

Page 24: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

17

Histogram based segmentation of Image of Handwritten text

An "image histogram" is a type of histogram that acts as a graphical representation of the tonal distribution in a digital image. It plots the number of pixels for each tonal value. Image histogram shows frequency of pixels intensity values.

Page 25: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Image Scaling

• Image scaling is the process of resizing a digital image.•Methods of scaling

–Nearest neighbor interpolation–Bilinear Interpolation–Supersampling–Vectorization

Page 26: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

18

FODEX and other systems

Other systems

•Handwriting University’s Handwriting Wizard.

Web-based computerized handwriting analysis system. The system consists of 9 handwriting characteristics. Each characteristic consists of one or more questions. Platform independent. It is a free handwriting analysis system. Text-based report type.

•Sheila Lowe’s Handwriting Analyzer.

Applying Gestalt or holistic graphology concept. Handwriting characteristics such as line spacing, margin, slant and pressure are represented by categories. At least 10 categories must be completed to generate a report. A standalone system. Runs on Windows. 04/15/2023

Page 27: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

•Advantage of FODEX over other systems–User Friendly–Faster–Automated–No other input from User beyond their Handwriting–More Accurate

19

Page 28: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

20

Technical Details

•JDK 1.7 Open Source Libraries for Image processing ImageJ, Fiji, Commons Imaging, Endrov, LeadTools , OpenCv, etc

•Java Web Start

•MySQL

04/15/2023

Page 29: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

UML DIAGRAM

21

Page 30: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

22

Page 31: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Mathematical Model.

Let S (be a main set of) ≡ {I, Isp, Igs, Ib, Itd, Ibt, Ith, Isg, Isc, If, Iout, DS}•Where,•S is the system.•I is the image•DS is Data Set•Isp is RGB splitting of image•Isp=RGBSplitting(I)•[sepB= RGBcolor AND 0xff•sepG= (RGBcolor >>8) AND 0xff•sepR=(RGBcolor>>16) AND 0xff]• 

23

Page 32: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Continue..

Igs=Grayscale(Isp)

[Igs=(sepR+sepG+sepB)/3]

 

Ib=GaussianBlur(Igs)

 

Itd=Threshold(Ib)

[Threshold th=128

If(gs>th) then gs=1 else gs=0]

 

Ibt=Binary(Itd)

 

Ith=Thinning(Ibt)

Isg=Segmnetation(Ith)

Isg={s1,s2,s3,s4…..}

Isc=Scaling(Isg)

If=FeaureExtraction(Isc)

If= {f1,f2,f3,f4,…..}

 

Iout=Comparison_Dataset(If)

Iout={Email, Log, PDF}

  24

Page 33: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

PROJECT PLAN

13

Page 34: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

IMPLEMENTATION TILL DATE

Page 35: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Page 36: forensic document examiner using graphology science
Page 37: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Page 38: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Page 39: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Page 40: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

CONCLUSION

We present image processing technique which is used to generate user’s profile. This profile will be matched with the traits of the graphology which are already defined. The system accepts scanned image from user and convert it to a form from which features could be extracted and profile can be generated. After all this processing the final output is generated and provided to the user in form of print or email

Page 41: forensic document examiner using graphology science

SKNCOE, Department of Computer Engineering 2014-15

Bibliography [1] Omar Santana, Carlos M. Travieso, Jesus B. Alonso, Miguel A. Ferrer Dpto. de Senates y COllUlnicaclones, Universldad de Las Palmas de Gran Canaria , “Writer Identification Based on Graphology Techniques”

[2] Perut Boribalburephan and Boonnatee Sakboonyarat , “An Algorithm Development for Handwritten Character Recognition by Personal Handwriting Identity Analysis [PHIA]”, Knowledge and Smart Technology(KST),2012 TH International Conference, Pages 6-10.

[3] Champa H N,K R AnandaKumar, “Artificial Neural Network for human behaviour prediction through Handwriting Analysis”INTERNATIONAL journal of computer application may 2010.

[4] Tristan Wright, “Handwriting Recognition with Artificial Neural Networks and Open CV”.

[5] Sargur N. Srihari, Ph.D.; Sung-Hyuk Cha, Ph.D.; Hina Arora, M.E.; and Sangjik Lee,M.S , “Individuality of Handwriting”

[6] Nobuyuki Otsu (1979). "A threshold selection method from gray-level histograms". IEEE Trans. Sys., Man., Cyber. 9 (1): 62–66. 10.1109/TSMC.1979.4310076

[7] Vassilis Katsouros and Vassilis Papavassiliou Institute for Language and Speech Processing/R.C. “Athena” Greece , “Segmentation of Handwritten Document Images into Text Lines”

Page 42: forensic document examiner using graphology science

Thank you ☺